There is increasing empirical evidence suggesting that a significant subset of eating disordered individuals with high degrees of trait impulsivity display complicated psychopathology, negative course of their eating disorder, and do not respond well to traditional eating disorder treatments. Very little data is available about the day-to-day activities and behavioral patterns of these individuals, frequently referred to as "mu1ti-impulsive" or borderline-type eating disordered patients. Furthermore, there is apparently no effective treatment for these individuals. This grant represents a conceptual model of self-destructive behavior in eating disordered individuals which incorporates both trait-constructs and daily events, moods, and behaviors to account for such self-destructive behavior. This model specifically examines the role of certain personality traits (affective lability, impulsivity), childhood adversity, daily events, daily moods, and daily coping behaviors in the prediction of self-destructive behavior. Furthermore, the present study proposes to test this model utilizing electronic daily diary technology to assess the behaviors of interest. We believe that this methodology will provide reliable data that is less affected by retrospective recall bias which characterizes previous research in this area. One hundred and twenty subjects will complete electronic diaries for thirty consecutive days, which will allow the collection of a sufficient amount of data to adequately test the proposed model. We believe that this methodology will allow us to identify antecedent and consequent conditions surrounding self-destructive behavior in eating disordered individuals. This will provide important information regarding the development of treatments for this group of individuals.